Let's Make Robots!

Gyroscopic Precession Robot (Version 2)

Shuffles forward via precession of the flywheel, sometimes picking its feet up.

Update 4/23/2014: Now radio controlled stepping, ESC on rotor motor makes for a high stepping robot.

 

This is a very crude implementation of John W Jameson's Walking Gyro that he patented in 1981. I had made a simple gyroscope (flywheel) using a dc motor and cd and was looking for something to use it in and I stumbled across Jameson's robot at this website: http://cyberneticzoo.com/walking-machines/1981-the-walking-gyro-john-w-jameson-american/

In his patent he details how it works including all the math.

There were some commercial toys produced from his patent, one even had directional ability but I could find no videos or further information on them.

Here is a photo of his robot:

 

 

Here is a photo of his patent picture:

Here is a photo of a fantastic motorized Mechano version:  I sure would like to have this one. The motor is on the other side of the flywheel.

My robot (that I named Bojangles cause he just kind of shuffles around mostly) doesn't use a pully arangement to tilt the gyroscope out of the vertical plane but rather a separate geared motor. Unfortunately the gearing is too slow so that Bojangles spends too much time turning and sometimes turns all the way around on one foot. Also, there is so much play in the gearing that it is unpredictable how much time will be spent making a step.. Also the mass of the flywheel affects it's movement, a small flywheel weight will cause it to shuffle whereas a larger one will make it pick it's feet up - also affected by the speed of the flywheel. Lots of variables that I couldn't change because my gearing was fixed as well as the length of the crank arm (too large causing too much tilt). My geared motor is about 6 rpm (ok for shuffling with a slower flywheel speed) but I'm guessing I need about 15 rpm to go straight while picking it's feet up and using a higher flywheel speed.

The feet should be able to rotate also whereas mine were mostly straight rods of fiberglass fishing pole.

The video does show mine picking his feet up (using two cds for a flywheel and a higher flywheel speed) but it's pretty precarious because the step time is too long and I believe the angle that I tilt the flywheel out of the vertical is too great. 

My robot shuffles mostly because my gyroscope flywheel (cds) are so out of balance. Operates kind of like a bristle-bot but the precession still moves the feet forward.

Here are some closeup photos of Bojangles:

Here he has one foot off the table.

 

This is the arm for the feet (missing) attached to the geared motor.

\

 

Photo of backside of geared motor:

 

I am using two cds hotglued together for the flywheel.

The geared motor's crank arm is a little too long so it swings the gyro out of the vertical a little too much.

I use two sets of batteries, one for the gyroscope and one for the geared motor but they could be both run off the same set. It was just easier using two sets so that I could change up the speed of the gyroscope without affecting the speed of the geared motor by using a different number of batteries.

Center of gravity needs to be over the pivot point of the arm to which the feet are attached. My robot has everything so spreadout that it is difficult to balance it and sometimes if the precession torque is too great he will fall over.

 

Here I used airplane wheels for his feet so that they would rotate when he moved forward but they made balancing difficult because they were so loose. 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

I have to say, I didn't understand what you were trying to do when viewing this previously.  Seeing the movie and the added pictures, now I get it!  

Very cool and this gets me thinking about the possibilities.  Thanks for posting it.

Regards,

Bill 

 

 

Yes its kind of a weird way to get motion and not too obvious if you haven't played much with gyroscopes, which I haven't until now. I think Jameson's use of precession to get forward motion is brilliant but not too useful except as novelty. But to watch the bot take a step and rotate on one leg is pretty cool. 

On version 2 I used a PowerBall rotor (can get them really cheap here in china) which spins up nicely and has a nice sound too. I had to have our university machine shop make an axle collar to connect the rotor to the motor. Using radio control allows control of motor speed and the stepping through a servo to pick the legs up rather than just using cyclical motor to control the legs.

Hope you try a gyro.

I have a bunch of brushless motors I intended to use with a quadcopter. They're outrunners so they have a pretty good about out rotational inertia without adding a disk. I'm very tempted to see how one would work with this sort of robot.

As others have said, this is an interesting project. Thanks for posting it.

I fly wings and have several burned out motors (I dont rewind). Will have to try too.

I totally love this project!

Gyros FTW!

I wonder what other 'strange looking' moves one could pull off using gyros to make 'strange' balance acts.

Hmm.. would there be any easy 'off the shelf' gyros that one could use? Encapsulated and all? Hard drives?! Hmm.. Would want to keep shaking & power useage to a minimum.

It's very hard to Google, as the word Gyroscope has so many articles. 'Encapsulated fast spinning wheel'?

Maybe roller skate wheels going really fast would do? But we want 'large thin wheels only heavy on the outside'.. Record players - nah - Basically we are looking for 'Flywheels'. I think. Have not managed to find a good solution. Hmm..

 

What would the behavior of walking robot be if it had a gyro in each knee? Would that be sufficent to keep it's balance? How would multiple gyros in the same plane and different planes interact? Lots of questions.

I really don't know enough about gyros to speak intelligently but I just feel they could be used a lot more.

It does seem that there are only two kinds commercially, the simple pull a string one and a lab quality monster with a huge motor on it, nothing inbetween. I guess machining ones own is an option for some. Also how small can you make one that is still effective. 

RPM is definitely worth investigating as there seems to be a sweetspot on my robot where he will pick up his feet and its not necessarily at the highest rpm.

 

I am definitely thinking about trying this out. This idea appeals to me so much.

As a first attempt, I might try putting apart, motorising, and encapsulating a toy gyro similar to this project?

http://makezine.com/projects/gyrocar/

Also, I think yours looks great, so don't pick on him! :)

I wish I had decent tools, dont even have a drill press. I teach ESL in a foreign country and hand tools are all I have to work with. I am thinking about commissioning some students here at the University where there is an Numerical Control lab to build one for me.